Longer-term health and social care strategies for stroke survivors and their carers: the LoTS2Care research programme including cluster feasibility RCT

Study design

This was a qualitative study involving semistructured interviews and a thematic approach to data analysis.

Participants

Participants were community-dwelling stroke survivors and their carers at two different time points: 9–12 months post stroke and between 2 and 4 years post stroke. This allowed for the identification of ongoing needs and provided an opportunity to reflect on what information and support had been useful and how this could be improved.

Potential participants were identified from an established research database held at Bradford Teaching Hospitals NHS Foundation Trust. In brief, stroke survivors were eligible for the study if they had a confirmed primary diagnosis of new stroke, were > 9 months post stroke, resided in the community and were able to provide informed consent (or consultee assent). Carers were eligible if they were identified by the stroke survivor as the main informal carer who provides support a minimum of once per week.

A maximum variation purposive sampling strategy was used to select a heterogeneous population with a range of disability levels (assessed by the Barthel Index19), high and low levels of unmet need,20 differing socioeconomic status (assessed by postcode), living circumstances (alone/with carer) and age range at the two time points.

Semistructured interviews

Unmet needs and the barriers and facilitators experienced in trying to overcome these were explored across all of the interviews (topic guides are provided in Report Supplementary Material 2).

Efforts were made by the researchers to tailor the interviews for stroke survivors with communication difficulties (by using pictures/adapting topic guides/use of keywords). The interviews were audio-recorded and transcribed verbatim. A thematic approach21 to data analysis was taken, with the transcripts analysed in two categories (those who were 9–12 months post stroke and those who were > 24 months post stroke), to establish any differences in the experiences at different time periods post stroke.

For the purpose of this analysis, needs were defined as those that stroke survivors perceived as challenges to overcome or address. Standard approaches to demonstrating trustworthiness and quality in qualitative research were used. 22 Throughout data collection and analysis, data, codes and emerging categories and theories were presented to and discussed with the research team and Programme Management Group (PMG) at regular intervals.

Identified needs, barriers and facilitators

The needs identified and the barriers and facilitators that the stroke survivors and their carers experienced are summarised in Figure 2. This figure highlights where barriers and facilitators relate to more than one need. For example, ‘building a support network’ is a facilitator across nine of the 13 needs. Other common barriers and facilitators include lack/loss of support, stigma, acceptance as a process and creative problem-solving.

FIGURE 2.

Stroke survivor needs, barriers and facilitators.

Thirteen key needs were identified:

1. managing and coping with a major life event

2. gaining control

3. managing emotions

4. reconstruction of identity

5. doing everyday tasks around the house

6. working towards physical and functional improvement

7. managing hidden consequences of stroke (e.g. difficulties with concentration and processing of information, memory impairments and mood swings)

8. obtaining usable information

9. sustaining flexible support networks

10. engaging in meaningful activity

11. overcoming financial concerns

12. maintaining relationships

13. managing beyond the home.

Some differences were recognised in the two time frames. Feelings of hope in terms of the need to work towards physical functioning were more common amongst those who were 9–12 months post stroke. These stroke survivors also talked more often about establishing a cause for their stroke as part of understanding and making sense of their situation (linked to the need to gain control and facilitate this process). More of the participants who were at least 24 months post stroke talked about the process of reaching acceptance. This was important for addressing many of their needs, including doing everyday tasks around the house, managing and coping with the major life event, managing the hidden consequences of stroke, reconstruction of their identity and maintaining relationships. Aside from these slight differences at each time point post stroke, the identification and management of needs remained similar across the two time points.

Four of the unmet needs related to emotional needs: gaining control; managing and coping with a major life event; managing daily stressors and strains; and reconstruction of identity.

Development of the survey

Using the information collated in the LoTSCare programme of work, a detailed six-page survey tool was formulated that collected data on the community stroke team (CST): where they received referrals from, what services they offered and for how long, and what other services were available in their area. This survey was sent to two community teams for feedback, and then e-mailed to the 29 Community Stroke Services that participated in the LoTSCare trial. The survey was sent out in August 2013 and a reminder was sent at the beginning of October 2013; however, only eight (28%) of the surveys were returned. The research team and PMG concluded that this pilot survey was too long and too difficult to complete in Microsoft Word (Microsoft Corporation, Redmond, WA, USA) and that the method of contact was not effective.

At the same time that the survey was being prepared, the Sentinel Stroke National Audit Programme (SSNAP) started to plan collection of national data on the 6-month stroke reviews and the Stroke Association offered the research team data on its service provision across England as commissioned from April 2014. Consequently, it was agreed that the SSNAP data would be utilised to inform our understanding of objectives 1 and 2, and a simple one-page survey would be sent out to all CCGs to achieve objectives 3 and 4. The Stroke Association data would be used to complement the survey findings.

The survey asked how long stroke services were commissioned for, whether or not any other providers commissioned services in the longer term and whether or not an annual review was commissioned (see Appendix 4).

Survey participants

Survey administration

The survey (see Appendix 4, Figure 10) and a cover letter, signed by Anne Forster (lead investigator) and co-investigator Matthew Fay, were sent via e-mail on 18 December 2013. The e-mail to the CCGs was addressed to the lead for longer-term conditions/stroke. Completion of the survey could be via an electronic Word document or a SurveyMonkey^® (Palo Alto, CA, USA) link. A written reminder was sent to all CCGs, and an e-mail reminder was sent to all community teams 1 month later on 20 January 2014.

The same e-mail and copy of the survey was also sent to the NIHR SRN managers who were asked to forward the survey to local CSTs. The survey was also promoted at the UK Stroke Forum Exhibition Stand (December 2013), and completed then by a number of community teams.

Data synthesis

Survey data were entered into a database and the length of service provision was categorised as (1) up to 6 months post stroke, (2) up to 12 months post stroke or (3) > 12 months post stroke. This service consisted of stroke or neuro-specific community rehabilitation teams. Commissioning of general community rehabilitation teams was not accepted as provision of support to stroke survivors.

Provision of other services was taken from the free text in the survey and grouped according to the type of service: (1) Stroke Association or (2) other support, including exercise classes, stroke clubs, etc. Provision of Stroke Association services was verified by the Stroke Association data; when there were discrepancies, the Stroke Association data were taken as being correct.

Information on the provision of 6-month reviews was taken from the SSNAP data report, and also from information on the commissioning of 6-month reviews in the Stroke Association data.

Provision of annual reviews was collated from the survey. The survey did not distinguish between 12-month general practitioner (GP) health checks and annual health and social care reviews.

Service provision

Of 203 identified CCG areas, 116 (57%) responded to the survey (see Appendix 4, Figure 11, and Report Supplementary Material 5). Of these, 30% (35/116) provided a stroke or neuro-specific community rehabilitation team service up to 6 months post stroke, 40% (46/116) provided a service up to 12 months post stroke and 29% (34/116) reported a service beyond 12 months post stroke.

Stroke Association Life after Stroke services were commissioned in 74% of CCGs, providing support to stroke survivors up to 12 months post stroke. In the survey, other support, such as exercise groups and stroke clubs, was reported by 57% (66/116) of CCGs.

Six-month and annual reviews

The return rate for the SSNAP 6-month review data was low (only 22.8% of patient records had an answer in this section). Overall, only 15% (3360/22,273) of survivors received a 6-month review.

In the Stroke Association data, 12% of CCGs commissioned the Stroke Association to complete 6-month reviews, and a further 16% of CCGs commissioned NHS community services to complete the 6-month reviews.

In the survey, 39% (45/116) of CCGs commissioned an annual review for stroke survivors (including 12-month health checks completed by GPs and annual health and social care reviews).

Study design

Having identified different models of post-discharge support, we originally intended to gain insights, through focus groups, from CSTs that exemplified the three models. However, given the diversity of stroke service provision and the professional make-up of CSTs, additional focus groups were required to gain a fuller, more informed understanding of the barriers to and enablers of delivering longer-term support.

On review of the survey data, eight CSTs were purposively selected to be invited to participate in focus groups on the basis of their location and the pattern of post-discharge support they provided. CSTs were selected from different English regions, including both urban and rural districts. Initial contact and invitations were extended to a lead member of the CSTs identified from the survey data (i.e. stroke service co-ordinator, team leader, head of service, service manager or therapy consultant), who subsequently discussed the invitation with team members and invited fellow team members to participate. There were no set inclusion criteria for participation other than to include secondary care clinicians as well as community-based professionals who undertook clinical, organisational or managerial roles in supporting stroke survivors.

Research governance

Approval was obtained from the relevant NHS trusts’ research and development departments.

Focus groups

Focus group methodology was used to explore the views of CST members as to their understanding of stroke survivors’ longer-term (i.e. ≥ 6 months post stroke) unmet needs, barriers that inhibit CSTs from addressing them and enablers that support them to do so. Written informed consent was taken from all participants, who were assured that all data would be treated confidentially, and that all reporting would be anonymised.

The focus groups were facilitated by a researcher, with a fellow researcher taking notes. An identical topic guide was used for each focus group (see Report Supplementary Material 7). The staff mix supported an exploration of different perspectives and produced insights into the service group dynamic. All focus group discussions were audio-recorded and transcribed, and data were thematically analysed. 21

Data analysis

The researcher (AP) began by familiarising himself with the transcripts, and the contextual nature of the materials contained in them. Subsequently, the researcher engaged in an interrogative process of sifting carefully through the data and identifying initial codes, later grouped into themes, relating to barriers to longer-term support for stroke survivors, and enablers of longer-term support for stroke survivors. Ongoing analysis then refined the specifics of each theme. Further analysis of these themes uncovered distinct categories in the overarching contextual framework of each theme. Hence, the complexity of service providers’ perceptions of their needs, and the factors that presented barriers to their being met, could be captured and explored. Finally, a selection of compelling extracts was used to exemplify identified themes.

Similarly, practical examples of how identified service provider needs were being met in the context of current organisational practice were classified as ‘actual enablers’. In addition, practices that potentially enabled service providers’ needs being met were classified as ‘potential enablers’.

Perceived needs of stroke survivors

The key longer-term needs of stroke survivors were identified by focus group members, and are presented as a summary here and in Appendix 5:

• a meaningful role and sense of identity

• psychological support

• ongoing information and advice

• non-stroke-specific wider community engagement opportunities

• stroke-specific group engagement

• appropriate health and social care at home

• longer-term supported self-management

• support with personal relationships

• support with physical health needs

• employment and financial support

• effective transportation to facilitate access to services and activities

• social acceptance by society at large.

Barriers and enablers

The barriers generally fell into the following broad categories:

• deficits of skills and resources

• availability of training

• prevailing cultural systems and organisational processes in the NHS

• failure of multiagency partnership-working.

Enablers were often antithetical to the barriers and could be grouped into the following categories:

• creative in-house approaches to training and educational enhancement

• flexible operational, managerial and cultural approaches

• effective multiagency partnerships

• strong training and educational links with other agencies.

Barriers relating to resources and skills deficits were beyond the control of CSTs. These included the absence of sufficient funded psychologist posts in their teams, inadequate public transport systems that meant survivors found it difficult to access community facilities and unwillingness on the part of employers to support survivors in undertaking retraining in order for them to return to work. Similarly, a lack of nationally funded opportunities to support survivors and employers in maintaining employment, bureaucratic procedures concerning state benefits and the way in which mainstream community facilities, such as gyms, are not mindful of the needs of survivors were additional barriers that CSTs were unable to influence.

Cultural and organisational processes in the NHS created additional barriers. A target-driven, largely inflexible approach that placed adherence to the requirements of the SSNAP database over a clinically driven, person-centred approach created considerable barriers. The organisational shift from a traditional community rehabilitative approach to a short-term ESD-style intervention approach reduced the potential for longer-term support. Additional barriers rooted in NHS organisational systems included bureaucratic telephony processes that inhibited survivors from contacting CSTs. The orientation of NHS commissioning, with a perceived relatively low emphasis on funding rehabilitative support in the community, and the absence of evidence-based guidance on the components of a good longer-term support structure, represented further barriers.

The absence of effective multiagency partnerships lay at the heart of key barriers inhibiting longer-term support. Poor strategic relationships between the NHS and social services manifested itself in the lack of a shared vision for longer-term support, with social services seen to be pursuing a narrower, more limited goal centred on social care packages in the home. High staff turnover, insufficient time spent in survivors’ homes and, on occasion, delivery of incorrect advice to survivors characterised the perceived social care agency approaches. Largely piecemeal joint budgetary arrangements, and a reluctance to engage with voluntary-sector bodies, provided further evidence of ineffective multiagency partnership-working.

Shortfalls in training involved both in-house training and training issues pertaining to other agencies. A lack of in-house training in CSTs meant that the capacity to support survivors affected by psychological difficulties was diminished. Staffing pressures and organisational patterns of work made it very difficult to deliver training to private care agency staff. GPs were felt to have inadequate training in spasticity management, and were poor at signposting survivors to other services. These training issues collectively reduced longer-term support for survivors.

Creative in-house promotion of education and training was delivered in various formats, and helped to enable longer-term support. In-house psychology training programmes to promote basic understanding of techniques to support survivors affected by low mood, and multidisciplinary team (MDT) clinics to promote a culture of interdisciplinary learning across professional boundaries were examples of enablers of longer-term support. Similarly, a culture that supports work-based learning and self-study, and the development of educational programmes for survivors that involve the participation of different agencies (e.g. therapists, benefit advisers, pharmacists), can also advance longer-term support in creating greater awareness among survivors on how to cope more successfully. Self-management programmes, such as Bridges,99 and the delivery of the Expert Patient Programme100 in a group environment were also cited as helping to achieve this goal. The delivery of peer support by survivors, who had undergone rigorous training, to offer guidance to survivors was also highlighted, in addition to the potential value of a keyworker approach in acting as a mentor/guide to more recent survivors.

Flexible operational and managerial approaches to delivering health care to survivors also seemed to enable longer-term support. Hence, an open-door referral policy that facilitated ease of access to therapists, collective working patterns within CSTs that ensured that the geographical region was adequately covered and the use of MDT clinics were indicative of flexible operational systems. Interdisciplinary methods that enabled more timely fitting of necessary equipment aids in survivors’ homes and a more clinically driven approach for delivering short bursts of therapy according to survivors’ needs further illustrated a flexible managerial approach that advanced longer-term support.

Although reported relatively infrequently, there were examples of multiagency partnerships that facilitated longer-term support. A post-stroke physiotherapy group, managed by leisure services and the local NHS, was an example of this. Furthermore, some services reported positive links with local employment services, and positive engagement with national programmes supporting disabled people into work. Community-based stroke groups, often run by the Stroke Association or other voluntary groups, sometimes involved different agencies coming together to support aphasic and younger survivors. A day centre that involved the participation of different voluntary groups, as well as a gym, and that hosted a cafe that promoted social interaction among survivors, was also cited.

Some support built around training and educational activities could also be effective. The positive impact of training of re-enablement workers by therapists to promote independent living at home could be considerable. Training survivors to utilise information technology (IT) methods to promote independence, particularly through using Skype™ (Microsoft Corporation, Redmond, WA, USA) and e-mail to enhance communication, was considered a positive approach to enhancing longer-term support.

Methods

With input from the RG and PMG, the principles for the intervention were developed based on those outlined in the grant proposal.

The total list of 23 needs, identified in WS1a from the qualitative work, literature reviews and helpline data and from the focus groups in WS2 (see Figure 5), was prioritised by members of our RG and our established CRAG. Members sorted cards of the needs into piles of ‘most important’ and ‘less important’ and provided explanations for their choices. The selection of needs were iteratively reviewed and refined in discussion with these groups.

Results

Principles for the intervention were developed, and included the overall aims of improving QoL and participation for stroke survivors; that the intervention be relevant and accessible to all stroke survivors and their carers; that it be responsive to context, be feasible and sustainable; and that it can be developed in a context of existing health and social care resources.

The RG and CRAG emphasised that the importance of specific needs would vary widely between individuals. However, they provided guidance as to the needs that our intervention should concentrate on, emphasising psychosocial needs and arguing that these were less likely to be addressed by current services, and, moreover, were most likely to influence QoL and participation.

The prioritised needs to address were gaining control, managing emotions, maintaining relationships, managing or coping with a major life event, reconstruction of identity, managing hidden consequences of stroke, communication, mobility, clinical problems, support around health and social care provision, obtaining usable information, sustaining flexible support networks and engaging in meaningful activity (see Appendix 6).

Of the 10 remaining needs from the list of 23, six were needs with existing service responses, and therefore remained as needs to be addressed by referral and not directly by the intervention (e.g. falls, driving, financial). Four needs had barriers and enablers similar to those already included, and were therefore likely to be encompassed in the intervention (e.g. ‘managing beyond the home’ has barriers and enablers that are similar to those of ‘engaging in meaningful activity’).

The RG discussed the following challenges:

• Complex psychological and relationship issues – it was agreed that these were beyond the scope of the intervention and should be managed through referral to existing psychological services and organisations (e.g. Relate).

• Carer-specific needs – although acknowledging the importance of recognising carers’ needs (and implementation of the Carer Act 2014101) it was agreed that the intervention would focus on survivors’ needs. Those delivering the intervention should be equipped to recognise problems and to help carers find alternative appropriate services. Detailed exploration of carer needs was not undertaken to inform an intervention. Carers’ Resource were an active participant in our research programme and ensured that we were aware of the needs and perspectives of carers and other family members at all times. Our research team had undertaken considerable previous work addressing carer needs. 45 A colleague in our unit undertook a linked PhD specifically continuing that work, focused on developing an intervention to reduce carers’ burden. 102

Methods

In accordance with the intervention mapping framework,103,104 behavioural outcomes for the 13 prioritised needs of survivors and their families were identified. For each of these behavioural outcomes, performance objectives were developed (what participants and agents in the environment need to do to achieve the behavioural outcome). Matrices of change objectives were developed, identifying specific determinants from the theoretical domains framework (TDF)105 that were likely to influence achievement of these performance (e.g. self-efficacy, skills) objectives in a range of domains. Fourteen domains of the refined TDF were used, along with two further domains particularly relevant to a stroke population: physical function and communication.

Results

A catalogue of 263 performance objectives and hundreds of change objectives designed to address the identified needs was developed (an example is provided in Appendix 7). As a result, the intervention mapping approach became too overwhelming in the face of the complexity, the interindividual variation and the lack of several specific behaviours that longer-term stroke survivors’ outcomes of QoL and participation could be focused on. The hundreds of change objectives that were generated highlighted the limited applicability of intervention mapping for an intervention designed to meet such a wide range of needs, each with a large variety of possible barriers and enablers. The working group agreed that there were too many performance and change objectives to take through the proposed steps of intervention mapping. In addition, the relevance of these objectives to overcoming the needs of survivors or carers would be very dependent on the particular needs and circumstances of the survivors and carers. There appeared to be some similarity between many of the performance objectives organised under different behavioural outcomes.

Instead of identifying methods and applications for each change objective in accordance with intervention mapping, the intervention plan was developed using:

• the similarities identified in the performance objectives above

• problem-structuring and shared knowledge creation, shaped by discussion with the PMG and RG

• information gained from all WSs and from the expertise of stroke survivors and carers

• consultation with health, social and academic professionals

• the research evidence review (see Appendix 8).

These sources converged to produce three intervention components: problem solving, building and sustaining a flexible support network, and obtaining usable information. When compared with the interim outputs from the intervention mapping work that had attempted to merge performance objectives across multiple needs, the approaches were compatible in cases for which it was feasible to check.

Recognised approaches to each component of the intervention were selected, consulting the relevant literature, the RG and the PMG. Practical applications to enact the components of the intervention were then considered and identified.

Problem-solving

To clarify, problem-solving was intended to address prioritisation of needs, and developing and enacting solutions that were specific and feasible for each survivor’s lived experience and circumstances. This would be allied with opportunities to access existing therapies when this was desired, deemed necessary and available. Existing approaches were adapted, because social problem-solving is a well-developed method with underpinning models, established components and approaches to application (e.g. problem-solving therapy106).

Building and sustaining a flexible support network

Building and sustaining a flexible support network enabled current problem-solving, but was also intended to provide a protective effect against the negative consequences of future challenges. Relevant methods and applications to incorporate into the intervention were identified.

Obtaining usable information

Access to usable information would also enable problem-solving, as well as satisfy the need to understand the personal and mutual ramifications of stroke and future prospects for recovery. In specifying this component, the RG and CRAG recognised what the evidence base suggested: that provision of large swathes of material would tend to overwhelm, rather than support, survivors, and that assistance in identifying and interpreting relevant information would better fit the principles of self-management.

Methods

The literature was searched for existing self-management interventions in stroke and other chronic conditions (details of the search strategy can be found in Report Supplementary Material 8). Existing materials were reviewed to determine whether or not they fitted with the selected approaches. When this was the case, these were adapted to the target group. When this was not the case, new materials were developed in accordance with the selected approach. The RG, PMG and CRAG contributed to this process.

Results

The comprehensive search of the literature relating to self-management in stroke produced 2492 titles. From these, 171 papers were scrutinised in detail. The form and content of numerous intervention materials were considered, in particular the Chronic Disease Self-Management Programme; the West Lothian Stroke Workbook; a problem-solving approach previously developed by Professor Allan House for stroke/diabetes; the Greater Manchester CLAHRC Patient-Led Assessment for Network Support (PLANS) website; Professor Lewin’s The Heart Manual; and The Pain Management Plan, a stroke recovery video guide developed in New Zealand.

For the problem-solving section, materials previously developed by Professor Allan House for stroke/diabetes patients were selected and adapted with reference to the latest iteration of problem-solving therapy. Further materials to support building a sustainable support network were developed and added. Stroke Association literature was selected and adapted to provide essential information for stroke survivors to enable them to understand their stroke and its repercussions. A short list of key national contacts was also developed to provide opportunities for survivors and carers to access relevant information. Furthermore, work was carried out with the Greater Manchester CLAHRC in adapting its PLANS website as an information resource identifying local activities and services. These materials were brought together in a booklet format, which was branded the ‘Moving Forward Guide’.

The Longer-term Unmet Needs after Stroke (LUNS) tool,20,107 originally designed to elicit needs for which there was a clear service response, was extended to incorporate the full range of needs identified in earlier work (LUNS+). The original intention to send this tool out to survivors in advance was considered, but, in discussion with the CRAG, PMG and PSC, it was decided that to do so would be overwhelming or perceived as negative by some survivors. From the extended LUNS tool, a priming tool was developed to explain the purpose of the meeting between survivor and site facilitator, including a sample of life domains that might be of interest and to provide space to record important changes since the stroke and challenges overcome.

Action groups

An action research, service improvement approach was employed to engage staff in a cyclical process to preliminarily test and refine the care strategy. Action groups were established at each site and the action research process involved input or dialogue, action, transformation, review and refinement. Each site was asked to identify facilitators who would be trained to lead local implementation (henceforth called the ‘site facilitator’). Each action group meeting was facilitated by a researcher from the central research team (henceforth called the ‘action research facilitator’) (further details are provided in Appendix 10).

Semistructured interviews

Semistructured interviews were undertaken with stroke survivors and their carers (when applicable) to obtain insights of their experiences of the intervention and their views regarding the care strategy. Interviews were conducted in survivors’ homes using a study-specific topic guide. Interviews were audio-recorded and transcribed verbatim. Field notes documenting contextual factors were collected by the researchers to aid analysis.

Semistructured exit interviews were also undertaken with the pilot site facilitators to explore their experiences and thoughts with regard to delivering different elements of the care strategy.

Expert panel

An independent panel of specialist clinicians, who would be available to participate in supervision and case discussion if required, was established. The idea of using an expert panel arose owing to concerns that the intervention might fail if stroke survivors’ problems were too difficult for the site facilitators to address and/or required specialist services (e.g. psychology and pain clinics). Site facilitators were instructed to submit anonymised case studies of any patients or problems they were having difficulty with to the independent panel, which would review the case and determine if it required very specialist input or if, with guidance, the facilitator could have addressed the problem. This process was introduced to enable the feasibility of the intervention to be assessed (i.e. would the majority of stroke survivors actually require specialist input?).

Focus groups and expert seminars

Focus groups comprising staff from the pilot sites and expert seminars with national experts were held at the end of the study to draw together the experiences and lessons learnt to refine the intervention and delivery mechanisms.

Challenges of the action research process

Challenges to the action research process included disagreement within action groups, challenges of moving forward without consensus, relative power of the action groups to engineer change given existing service frameworks and patterns of working, and differences in autonomy/freedom from established ways of working.

Number of facilitators

Although the action group provided a supportive structure, in two of the three sites, one person was tasked with delivering the intervention. It was recognised that this was not ideal as information provided from the sites indicated that the associated workload would preclude all stroke survivors being seen. Therefore, for the feasibility trial, it was emphasised that at least two people should be trained in the new approach.

Identification of stroke survivors

All three services established mechanisms to identify stroke survivors at the appropriate time point. However, at one of the sites, the facilitator undertook a detailed audit to check the throughput and noted that a number of stroke survivors had been missed off of the referral lists for no apparent reason. ‘Talking through’ the care pathway as it currently exists might highlight gaps and duplication.

Timing of delivery

Timing of delivery of the intervention was confirmed to co-ordinate with or act in place of the recommended 6-month review.

Building a relationship with the stroke survivor

Members of the action groups and stroke survivors, in their interviews, both emphasised the importance of building a rapport in order to successfully deliver the intervention.

Dealing with specific stroke survivor problems

The facilitators requested that training should include a focus on dealing with people with anxiety and depression, as that had proved challenging on occasions.

Expert panel

The research team had concerns that problems might arise that were (1) outside the expertise of the facilitators or (2) required specialist input. The availability of the expert panel was highlighted through the action groups, but no cases emerged. This provided reassurance that addressing the identified unmet needs of stroke survivors was within the remit of these facilitators.

Screening data

Screening data were collected by research staff for all stroke survivors identified as potentially eligible, and included basic demographic details (age, sex, ethnicity), dates of hospital admission and discharge, National Institutes of Health Stroke Scale (NIHSS) score at admission, modified Rankin Scale (mRS) score at discharge and availability of a carer. If applicable, reasons for ineligibility and reasons for declining participation were also recorded.

Baseline data

For those stroke survivors registered to take part in the trial, the following baseline data were collected by research staff: ethnicity, date of birth, sex, NHS identifier (ID), living arrangements (i.e. alone/with another person), carer identified by stroke survivor, address and telephone number, GP details, date of stroke, stroke severity (e.g. NIHSS score at time of admission), mRS score at discharge, level of impairment at recruitment (e.g. speech impairment, difficulties with communication) and preferred method(s) of contact. Stroke survivor baseline assessments were administered by research staff in person or by post (according to stroke survivor preference).

Follow-up data

The central trial team followed up stroke survivors via postal questionnaires at 3, 6 and 9 months post recruitment, supported by postal, telephone and text reminders if questionnaires were not returned in 2 weeks. Proxy completion of questionnaires was permitted. If outcome measures were not returned by post, telephone interviews were conducted to maximise data collection.

The questionnaires included:

• the World Health Organization Disability Assessment Schedule (WHODAS) 2.0,122–124 36-item version

• the Warwick–Edinburgh Mental Well-being Scale (WEMWBS)125–128

• the EuroQoL-5 Dimensions, five-level version (EQ-5D-5L)129–132

• the ICEpop CAPability measure for Adults (ICECAP-A)133–137

• the 13-item Short Form Patient Activation Measure (PAM) survey116,117

• the LUNS tool20

• relevant questions adapted from the Millennium Survey of Poverty and Social Exclusion138 and GP Patient Survey. 139

The questionnaires included a single overall life satisfaction question and a request for information about who completed the pack (stroke survivor/proxy) and how much help was provided.

Health, social care and voluntary- or third-sector service use was collected, together with costs, from participants using a resource use questionnaire.

The possibility of using NHS Digital data to provide information on stroke survivors’ hospital admissions and outpatient attendance during the trial was also explored.

Stroke services

Fifty stroke services responded to the initial contact with an expression of interest, of which 25 (50% of interested) returned feasibility forms, 13 (26% of interested) entered site set-up, 10 (20% of interested) consented to participating in the trial, and all 10 sites were randomised. During screening and recruitment, 14 services (28% of interested) were found to be ineligible, and 26 services (52% of interested) lost interest or ceased contact.

Five services were randomised to each arm, with minimisation characteristics balanced across arms (see Appendix 15, Table 10). Other site characteristics show the range of service provision across sites (Table 5).

Participants and carers

Of 1127 stroke survivors who received care across the 10 services and were screened for participation, 1034 (91.7%) were eligible, 367 were interested (32.6% of screened, 35.5% of eligible) and 269 were registered to participate in the study (23.9% of screened, 26.0% of eligible; see Figure 7). There was variation across services in the proportion of eligible participants who were registered (7.7–34.5%) (see Appendix 15, Table 11): it was slightly lower in the New Start services [145/587 (24.7%)] than in the usual care services [124/447 (27.7%)] (see Appendix 15, Table 12). The number of participants recruited per site ranged from 4 to 89 (usual care, 10–42; New Start, 4–89), with the mean number of participants per site being higher in the intervention arm (usual care, 24.8; New Start, 29).

Eighty-five carers of registered participants were also recruited to the study: 39 in the usual care arm and 46 in the New Start arm.

Comparison of recruitment rates with Sentinel Stroke National Audit Programme data

A table comparing SSNAP-reported figures with recruitment and clinical screening figures can be found in Appendix 16.

Facilitator training and competency assessment

Intervention training comprised an initial 2-day training session and two follow-up sessions. Facilitators then had the opportunity to practise intervention delivery before the trial commenced and were asked to complete two structured reflective reports focusing on New Start delivery during this practice phase. Training was adhered to fairly well, with only a small number of facilitators not completing or attending all aspects of training (see Appendix 15, Table 19). Facilitators were appraised regarding their competency in delivering the New Start intervention. All were deemed competent (see Appendix 15, Table 20).

The New Start intervention was delivered via face-to-face meetings with the stroke survivor. Overall, 86 out of 145 (59.3%) intervention trial participants had at least one intervention meeting. These participants attended an average of 1.14 meetings, each lasting approximately 1 hour (see Appendix 15, Table 21). Twelve of the 15 trained facilitators went on to deliver the intervention to registered participants, with a high degree of variability in the number of patients seen (range 1–24), number of visits (range 1–25), average number of visits per participant (range 1–2) and average duration of visit (range 20–82 minutes) (see Appendix 15, Table 23).

Baseline characteristics were compared between trial participants who did and trial participants who did not receive the New Start intervention. Notable differences included levels of higher education (49.2% among those who did not receive New Start vs. 39.5% among those who did receive New Start) and in the number of unmet needs, which was lower in those not receiving the intervention (see Appendix 15, Table 25). Those receiving the intervention had, on average, a longer stay in hospital as a result of their stroke (13 vs. 8 days). There were also differences apparent in the levels of stroke severity between these groups (measured via mRS and NIHSS) and in the level of language ability after stroke, but large numbers of missing data for these measures make interpretation difficult.

New Start sessions: non-trial participants

In keeping with the cluster trial design, all stroke survivors at intervention sites were offered the intervention, not all of whom consented to outcome data collection. Of the 442 non-study participants contacted, 294 (66.5%) had at least one intervention meeting (see Appendix 15, Table 22), which is slightly higher than the rate for study participants (59.3%). This population of stroke survivors attended, on average, a similar number of meetings as study participants (1.16), but these were, on average, shorter, at around 52 minutes. There was, again, a high degree of variability in the number of patients seen (range 3–43), the number of visits (range 4–47), the average number of visits per participant (range 1–2.33) and the average duration of a visit (range 31–84 minutes) for each of the 13 facilitators who delivered the intervention to non-study participants (see Appendix 15, Table 24).

Usual care

Only four of the five usual care sites offered a 6-month review, compared with all five intervention sites (see Appendix 15, Table 26). Across usual care sites, 86% of stroke survivors who were offered stroke care between 6 and 12 months post stroke were seen or spoken to; the majority (87.1%) of these had one contact, with an average duration of contact of 54 minutes (see Appendix 15, Table 27). Variation in usual care was observed across study sites (see Appendix 15, Table 28).

Uptake of reviews: New Start and usual care

The uptake of 6-month reviews across all services was 58.7%; however, it varied widely, from 9.7% to 100% (see Appendix 17).

A telephone invitation with an opt-in review was the most common approach to offer; however, a letter of invitation with a pre-booked appointment (opt out) resulted in the highest levels of uptake, on average.

Home was the most common location of review delivery, and resulted in higher levels of uptake, on average.

Unblinding

There were 14 reported occasions of New Start facilitator unblinding to research participation, occurring across three sites and 13 patients (see Appendix 15, Table 29). Most unblindings occurred within the first month post registration, an average of 20 days after registration (see Appendix 15, Table 30 and Figure 12).

Response rates to self-reported questionnaires

Return rates of patient-completed questionnaires were assessed by arm at all time points. Return rates were consistently higher in the New Start arm than in the usual care arm for all questionnaires at all time points (see Appendix 15, Table 31). This difference was particularly visible in the return rates of the WHODAS, WEMWBS and LUNS tool questionnaires at 9 months (usual care, 81.5–84.3%; New Start, 88.5–91.6%).

Completion rates of outcomes

The WHODAS and WEMWBS questionnaires were administered at baseline and at 6 and 9 months. There was a higher proportion of partially completed questionnaires at 9 months for the WHODAS, a 36-item questionnaire, than for the 14-item WEMWBS (66.9% vs. 10.4% of available participants) (see Appendix 15, Table 32). However, the flexibility of the WHODAS simple score (it can be reduced to 32-items for those not working) meant that, at 9 months, the proportion of partially completed questionnaires that could not be scored was higher for the WEMWBS than for the WHODAS with simple scoring (44.0% vs. 20.0%). This was not the case for the more rigid complex scoring of the WHODAS, for which 50.0% of partially completed questionnaires were unscored because of missing items. At each time point for both questionnaires, return rates in the New Start arm were slightly higher than those in the usual care arm.

The PAM was administered at baseline and at 3 and 6 months. Of the partially completed PAM questionnaires at 6 months, 47.6% (20/42) were unscored because of missing items.

Study within a trial assessing method of administration

At 6 months, 126 participants were randomised to receive a single booklet and 128 participants were randomised to receive two shorter booklets – 254 participants in total; one of the participants randomised to receive a single booklet withdrew shortly after randomisation, hence 253 participants in total were sent booklets. At 9 months, after accounting for deaths and withdrawals between 6 and 9 months, 116 participants (92.1% of those randomised) were sent a single booklet and 123 participants (96.1% of those randomised) were sent two shorter booklets.

At 9 months, participants were more likely to return follow-up questionnaires when these were administered as a single booklet than as two shorter booklets. At 9 months, 215 participants returned the booklets by post (90.0% of those sent). A total of 108 participants returned the single booklet (93.1% of those sent); however, among those sent two booklets, only 105 returned both (85.4% of those sent) and two returned a single booklet containing the outcome measures only (1.6% of those sent).

For those returning questionnaires at 9 months, item completion rates were also affected by questionnaire format. Health-care resource use alone and the full battery of measures (i.e. all health outcomes and resource use) were more likely to be completed in the single-booklet arm than in the two-booklets arm. Completion rates of health-care resource use alone in the single-booklet arm and the two-booklets arm were 98.1% and 89.7%, respectively (91.4% and 78.0% of those sent, respectively); full outcomes completion rates were 97.2% and 87.9%, respectively (90.5% and 76.4% of those sent, respectively) (see Appendix 15, Table 33).

Similar trends were also observed at 6 months.

Carers were also more likely to return the questionnaires when sent a single booklet. At 9 months, 30 carers returned the single booklet (88.2% of those sent), whereas 30 returned both (81.1% of those sent). For those returning questionnaires, there were no differences in item completion rates between groups at 9 months and a very small difference between groups at 6 months; however, when assessed as a proportion of those sent questionnaires, all types of outcomes at both time points were more likely to be complete for those sent a single booklet. The number of carers available and assessed in this analysis is low; thus, this finding should be interpreted more cautiously than that for stroke survivors (see Appendix 15, Table 34).

Overall, these results suggest that outcome follow-up rates are maximised when sending all required questionnaires in a single, longer booklet rather than when splitting them across two, shorter booklets.

Statistical outcomes

Questionnaire outcomes were measured at both patient (see Appendix 15, Table 35) and cluster level, with significance testing performed on the cluster-level point estimates of the WHODAS and WEMWBS questionnaires, at the 5%, 33% and 49% significance levels. The cluster-level estimates in the intervention arm used data from only four of the five intervention sites, as one site recruited only four participants and numbers with outcome data were insufficient for such analysis. At the cluster level, the WHODAS score with simple scoring was found to be significantly lower in the intervention arm than in the usual care arm at 6 months (mean difference 3.14, 67% CI 0.76 to 5.51) at the 33% significance level (see Appendix 15, Table 36). This finding was not observed at 9 months, and consistent significant differences were not observed at any other time point or for any other outcome.

The change in PAM levels between time points was assessed graphically via bubble plots (Figure 8). This shows, by arm, how many participants moved between PAM levels at baseline and at 3 and 6 months, and also summarises missing data. We can see that, between baseline and 6 months, 26.5% (30/113) of stroke survivors in the New Start arm with complete data improved their PAM level, compared with only 20% (17/85) of usual care participants. This pattern was also seen between 3 and 6 months [New Start, 28.8% (32/111); usual care, 26.8% (22/82)] and between baseline and 3 months [New Start, 23.4% (29/124); usual care, 20.0% (20/100)]. Participants in the usual care arm were also more likely to have missing PAM data than those in the New Start arm at all time points. The mean PAM scores in both groups (see Appendix 15, Table 35) did not indicate change in line with the reported 4-point minimal clinically important difference. 141

FIGURE 8.

Change in PAM level: (a) between baseline and 3 months; (b) between 3 and 6 months; and (c) between baseline and 6 months. Note that, in these plots, the numbers on both axes represent the PAM categorical level at each specified time point, with 1 being the lowest, least active level (individuals tend to be passive and feel overwhelmed by managing their own health. They may not understand their role in the care process) and 4 being the highest, most active level (individuals have adopted many of the behaviours needed to support their health, but may not be able to maintain them in the face of life stressors). Zero represents a missing value at the time point labelled on the respective axis. The size of, and number in, each bubble represent the number of patients in each group. The bubbles above the diagonal lines show patient groups whose PAM levels have improved over time.

Carer outcomes are summarised in Appendix 15, Table 37. The CBS was used to measure carer burden, with a higher score indicating a higher level of burden. At baseline, carers in the usual care arm had higher average CBS scores (usual care, 48.7 points; New Start, 45.6 points), as was the case at 3 months (usual care, 48.7 points; New Start, 47.7 points), but this reversed at 6 months (usual care, 43.7 points; New Start, 46.8 points) and 9 months (usual care, 44.9 points; New Start, 48.2 points) (see Appendix 15, Table 37). An explanation for this change is the relatively small sample size (85 carers) and large numbers of missing data, especially in the usual care arm. It is also thought that more severely affected stroke survivors, and thus more burdened carers, are less likely to be followed up as the trial progresses.

Sample size estimation

To inform the sample size estimation for a definitive trial, information is required on the average cluster size and coefficient of variation, as well as an estimate of the ICC for the proposed outcomes.

The mean cluster size, as defined by the number of registered participants per site, was 29 [standard deviation (SD) 34.18] in the intervention arm (145 registered participants; five sites) and 24.8 (SD 12.64) in the usual care arm (124 registered participants; five sites). These correspond to a coefficient of variation in cluster size of 1.18 in the intervention arm and 0.51 in the usual care arm. The high level of variance observed in the intervention arm is due to the inclusion of one small site (site 2a, four participants) and one large site (site 3, 89 participants). If the small site is removed, as was done in the cluster-level significance tests, the mean cluster size in the intervention arm changes to 35.3 (SD 36.02) (141 registered participants; four sites), and the coefficient of variation of cluster sizes changes to 1.02.

Intracluster correlation coefficients were calculated for the WHODAS (simple and complex scoring) and the WEMWBS questionnaires. They were calculated by treating recruiting sites as clusters, and evaluating the ICC for each questionnaire score at the given time points. Site 2a was omitted from the calculations as it contained only four participants. Because of the small number of clusters and participants, it was difficult to produce reliable ICC estimates for the outcome measures. The range of ICC values produced was between 0.01 and 0.02. Results are provided in Appendix 15, Table 38; cases for which the ICC estimate is zero do not have an accompanying 95% CI. Estimates for the coefficient of within-cluster variation were also produced with 95% CIs, to also aid in the sample size estimation for a definitive trial.

Recruitment

More than half of sites had recruitment periods of > 6 months, so, in addition to assessing the total number of stroke survivors recruited per site, a prorated figure is also evaluated against the progression criteria (see Appendix 15, Table 41).

The overall average number of recruited stroke survivors per site, prorated to a 6-month period, was 24.1, fulfilling the green requirements for this criterion (i.e. ≥ 20). Five of the 10 sites achieved green status, recruiting at least 20 patients when scaled to a 6-month period. Another four sites achieved amber status (≥ 10 patients); however, three of these fell within the ‘acceptable range’, recruiting at least 12 patients over a scaled 6-month period. One of the 10 recruiting sites was assessed as red in this criterion, as it failed to recruit an average of 10 patients over a 6-month period.

Follow-up

Follow-up is assessed via the number of questionnaire booklets received at 9 months post registration. Of the 269 patients in the trial, 216 booklets were returned at the 9-month time point, giving a follow-up rate of 80.3% and achieving green status for this criterion (see Appendix 15, Table 42). This includes all modes of administration, and both survivor- and proxy-completed booklets.

The follow-up rates are further split by treatment allocation, showing that follow-up rates were slightly higher in the New Start arm than in the usual care arm, with 84.1% of New Start participants returning the booklet, compared with 75.8% of usual care participants. Both groups achieved green status on this criterion, however, with follow-up rates of > 75%.

Intervention implementation and delivery

Intervention delivery is evaluated as the number of stroke survivors being offered at least one session of the intervention. When looking at data returned on the relevant case report forms, the proportion of stroke survivors offered the intervention varies significantly by site. Only three sites achieved the 75% proportion required for green status, with the overall proportion standing at 70%, placing it in the amber region (see Appendix 15, Table 43). However, there were issues with sites completing and returning these forms, and so these data are not believed to be complete or accurate.

Additional data were collected from the sites to see if more patients were offered the intervention than reported. When combining these data with those reported on forms, we see a much improved picture of intervention delivery, with all sites offering at least 75% of patients a session of the intervention, and with an overall proportion of 95.2%. This highlights a potential data collection burden on facilitators, which should be noted when planning a definitive trial.

In total, all five intervention sites had at least two facilitators deemed competent, delivered the New Start intervention and provided it to patients, securing green status for this criterion (see Appendix 15, Table 44). However, at some sites, there were concerns regarding the number of patients being offered, accepting and receiving the intervention.

Methods

Colleagues undertaking recruitment initially voiced concern at the feasibility of recruiting people predominately by post. However, the process progressed more smoothly than envisaged. No potential participants complained about being approached in this way. The research/service configurations were slightly different in each site. In one site, recruitment was undertaken by staff who were able to give potential participants ‘advance warning’ during their hospital stay that they would be contacting them in the future. However, this was not the case in the largest recruiting site (site 3).

Screening and baseline data collection was kept deliberately minimal and mirrored the information provided for the SSNAP (NIHSS and mRS scores). However, for some sites, staff from the research directorate did not have access (for appropriate information governance reasons) to the SSNAP record; therefore, these data are missing. For similar reasons, the recruiting staff were not always aware of any communication and cognition difficulties the stroke survivor may have, or the presence or absence of a carer. Postal recruitment influenced the entry criteria for the study. Our preference was to exclude stroke survivors in receipt of palliative care; however, we concluded that there was no appropriate way to frame that ‘qualifying’ question in a postal invitation.

Recruiting participants to longer-term stroke studies has previously been challenging, as there is no central register of stroke survivors and, once provision of services is ended, which might be soon after stroke for some, there are limited opportunities to identify and engage with this population. We believe that this study is one of the first to successfully recruit a cohort of post-stroke survivors by post. Colleagues in the recruitment teams found the process smooth and resource-efficient.

Rates

Participants were recruited over 8 months, which was shorter than envisaged in the original grant application, with considerable variation between the sites. One site recruited a large number of participants; however, it was also a large service with a higher throughput of stroke survivors. In one rural site, the throughput was much smaller; consequently, the rate of recruitment was lower. Recruitment remained low at one urban site (n = 4), despite site visits and involvement of the CRN. In this geographical area, one community service received referrals from three acute hospitals. The community service (site 2a) received referrals as usual during the study, providing the intervention to 85 individuals, and the process evaluation reported that the New Start intervention was delivered with a relatively high level of fidelity. Although one of the recruitment staff had a period of absence towards the end of the recruitment period, we were largely unable to identify why the clinical services were not able to identify stroke survivors and invite them to participate in the study.

The recruiters themselves suggested a number of different reasons for variability in uptake, including socioeconomic factors and the context provided by existing stroke service provision. Capacity of the recruiting staff negatively affected recruitment in only one site, where a large number of research projects (not in stroke) were being undertaken, thereby reducing the resources available for this study.

Dose

Appendix 20 presents a summary of New Start activity data, as recorded by facilitators. In most cases, New Start was not delivered entirely as intended:

• Although 69.7% (377/541, where 541 is based on the number of stroke survivors with completed New Start activity records) of stroke survivors approached during the delivery phase received New Start, 75% (n = 284/377) of these did not receive the minimum dose, defined a priori as an initial visit plus follow-up.

• New Start was often introduced in such a manner that the underlying purposes and expectations were unclear.

• Facilitators only briefly explored social networks, rather than rigorously mapping them to identify supporters for future action-planning; this potentially narrowed problem-solving options. There is little evidence that facilitators were able to help stroke survivors develop social networks or use them to address ongoing needs.

• Although observational data suggest that unmet needs were identified on many occasions, activity records indicate that collaborative goal-setting took place during 5% (n = 20) of cases.

Nevertheless, it should be noted that several facilitators did successfully deliver aspects of the intervention, such as goal-setting/action-planning, and one facilitator delivered the intervention with a high level of fidelity.

Stroke survivor response

Stroke survivors could be unclear about aspects of New Start, that is they assumed that they were receiving a standard follow-up appointment, rather than an intervention containing elements of self-management. Consequently, the ongoing needs they identified were generally related to their physical health. Although they could struggle to engage in problem-solving and goal-setting, and a minority changed their activities as a result, most stroke survivors reported benefiting from the intervention because they felt supported and understood. Occasionally, they appeared to fully understand the purposes of the New Start intervention and engaged in all aspects of the intervention as intended.

Measurement of outcomes, resource use and costs

Quality-adjusted life-years were calculated based on patient and carer health state utility values obtained from the EQ-5D-5L questionnaire at baseline and at 3, 6 and 9 months. 145

Information on all health-care resource use during the trial was collected using patient- and carer-completed questionnaires at 3, 6 and 9 months, and converted to costs using appropriate UK unit costs146,147 (see Appendix 21, Table 52). The cost of the intervention was estimated as the cost of the 6-month review meeting along with any associated follow-ups. Total costs for each patient were calculated as the sum of costs assigned from hospital, community health and social services and the intervention cost, along with out-of-pocket costs incurred by patients and their informal carers.

Adjusting for baseline imbalance

A multiple regression analysis was used to estimate differential mean QALYs and predict adjusted QALYs controlling for utility at baseline.

Missing data

When there were missing QoL or cost follow-up data, multiple imputation methods were used to generate estimates of missing values based on the distribution of observed data, as per recommended best practices for economic evaluation alongside clinical trials. 148

Cost-effectiveness analysis

The primary analysis consisted of a cost–utility analysis over the 9-month trial period and included adjustment for baseline variables and imputation of missing data. The incremental cost per QALY gained from the New Start intervention compared with usual care was calculated, producing an incremental cost-effectiveness ratio (ICER). 149

Sensitivity analyses were conducted to explore the impact of assumptions made in the primary analysis and alternative perspectives for analyses.

The level of sampling uncertainty around the ICER was explored using a non-parametric bootstrap to generate 10,000 estimates of incremental costs and benefits. This was used to illustrate the probability that the New Start intervention is cost-effective at a range of threshold values.

Resource use and costs

Resource use throughout the trial, broken down by item and associated costs, is presented in Appendix 21, Tables 53 and 54. Use of health-care services was higher in the New Start arm, but use of private health-care services (paid for out of pocket) was higher in the usual care arm. The multiple regression analysis indicated that the difference in observed costs between groups was not statistically significant (p > 0.05, 95% CI –899.646 to 3211.452).

Quality of life

Patient and carer EQ-5D-5L and ICECAP-A scores are presented in Appendix 21, Tables 55 and 56, respectively. There was little difference in EQ-5D-5L scores over the trial period in either arm, and the multiple regression analysis indicated that there was no significant difference in total QALYs gained between groups (p > 0.05, 95% CI –0.043 to 0.014).

Missing data

Complete and missing resource use and EQ-5D-5L data are presented in Appendix 21, Tables 57 and 58, respectively. EQ-5D-5L scores were complete at all follow-up points for 195 (72%) patients and 35 (41%) carers. Resource use questionnaires were complete at all follow-up points for 180 (67%) patients.

Cost-effectiveness results

Cost-effectiveness results are presented in Table 6. This indicates that New Start would not be the favourable option as the treatment for stroke survivors, as the QALY gain is lower in the New Start arm. However, if a whole health-care system approach is taken, the New Start intervention has the potential to be cost-effective in a resource allocation sense, given that New Start is less costly but not much less effective (small difference in QALYs), that is the money saved has the potential to generate more ‘health’ elsewhere in the system. 149 However, these results should be viewed with caution, as this is an exploratory analysis of feasibility data only.

Bootstrapped estimates of the incremental costs and effects indicated that the New Start intervention is unlikely to be a cost-effective use of resources (see Appendix 21, Figure 17). At a cost-effectiveness threshold of £20,000 per QALY gained, the New Start intervention showed a 48% probability of being cost-effective (see Appendix 21, Figure 18).

The results of the sensitivity analyses are presented in Appendix 21, Table 59. These results showed a great deal of variation in the cost-effectiveness estimates for each scenario explored, demonstrating substantial uncertainty around the results.

Model

The model (Figure 9) is based on QoL rather than clinical events. This approach was deemed appropriate for modelling the impact of the New Start intervention, as the outcomes related directly to changes in QoL. It does, however, create an unusual co-dependency between QoL and health states (because QoL defines both health state and QALYs gained). The model was populated using data from the trial to inform transition probabilities, health-state costs and utilities as treatment costs associated with New Start and usual care. For model parameters that could not be collected within the trial, including long-term mortality following stroke, recommended best practices for identifying and synthesising evidence from the literature were followed. Treatment costs for each arm were taken from the LoTS2Care trial data and represent the average cost in each arm for health-care consultations and visits associated with New Start or the usual follow-up care 6 months post stroke. Model parameters are presented in Appendix 21, Table 60.

FIGURE 9.

Lifetime decision-analytic model to compare New Start with usual care following stroke.

Validation, sensitivity analysis and value-of-information analysis

Model validation was conducted with reference to the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) best-practice guide for model transparency and validation. 150 As in the within-trial analysis, the incremental cost per QALY gained was estimated and compared with a cost-effectiveness threshold of £20,000 per QALY. Deterministic sensitivity analysis was used to explore possible alternative scenarios to those used in the base-case analysis. A value-of-information analysis was conducted to explore the costs associated with the uncertainty in the results, and the expected value of perfect information (EVPI), representing an upper bound on the value of conducting further research, was estimated.

A full description of the methods, including the parameter values, and the results is presented in Appendix 21.

Generating initial codes

A data-driven approach was adopted to avoid losing a contextual understanding of experience. Line-by-line coding was conducted on each transcript, using an active voice to ensure that codes remained close to the text.

Searching for themes

Emerging ideas were documented alongside the line-by-line codes to establish similarities and differences in each transcript. For each transcript, data segments from the initial coding were organised into meaningful groups, supported with memos. Thematic maps were created (one for each transcript) to establish inter-relationships between the codes and themes for each participant.

Reviewing themes

The themes identified in each transcript were considered in terms of whether or not they captured the essence of what was in them. Progression to across-case analysis involved looking across the data set from each interview participant to establish the similarities and differences in the experiences of stroke survivors in the longer-term post stroke. An overall thematic map was developed for each time point, documenting where themes inter-related. Documenting the research process in this way enhanced dependability and confirmability. 22

Defining and naming themes

As the thematic analysis progressed and the themes were refined, the focus was placed on ‘needs’ identified in the stroke survivors’ and caregivers’ narratives and the barriers to and facilitators of addressing needs. For the purpose of this analysis, needs were defined as those that the individual see as a challenge to overcome or address. These carry importance because of the meaning associated with them. In some cases, these needs were dyadic and co-constructed by the stroke survivors and the carers.

A secondary analysis was conducted that involved re-examining the themes for each participant and developing a series of diagrams to establish needs, and the barriers to and facilitators of addressing needs, at each of the two time points.

Standard approaches to demonstrating trustworthiness and quality in qualitative research were used, including the clear documentation of the research process (methods, analysis and any problems encountered and solutions found). The software program NVivo 10 was used to store, organise and code the data. Memos concerning coding and emerging themes and theories were also recorded throughout the analysis process. Throughout data collection and analysis, data, codes and emerging categories and theories were presented to and discussed with the Study Management Group and PMG at regular intervals. The emerging findings were also presented to stroke professionals, as well as to stroke survivors and their caregivers in the CRAG. Comments received were considered alongside the ongoing analysis.

Managing and coping with a major life event

The stroke had affected the survivors and their caregivers to some extent. Some were more affected initially, but quickly reached a point where they were continuing as they did previously, for example Jane and David. Others took longer to cope and manage with the effects of the stroke across different aspects of their lives, for example Carla. For the majority of the stroke survivors, the stroke was viewed as unexpected and shocking and caused some level of disruption to their own, and their family’s, lives, even temporarily. Managing and coping thereafter is a longer-term issue; therefore, stroke survivors and their caregivers needed to be able to find ways of managing and coping in terms of moving forward from this major life event.

Stroke survivors and their caregivers adopted many beneficial strategies (‘facilitators’) for coping with the ‘major life event’. Making comparisons with others was an interesting strategy as this can have both negative and positive consequences for stroke survivors and their caregivers. For some, making these comparisons enabled some realisation that they are lucky to be alive or that their difficulties are not as bad as they could be. This left them feeling fortunate, despite experiencing a major life event:

I was so delighted to be OK when I’d heard of people who are, you know, disabled or speech problems and all sorts of things that you hear after and I was so grateful that I was very happy, actually.

Julia

In contrast, some stroke survivors felt that they were not doing as well as they could be when they compared themselves with others. Marilyn felt this way when she compared herself to someone who had experienced a milder stroke:

Yeah, and I said to [husband] ‘oh, she’s walking already’ and he said ‘yeah, but her stroke weren’t as serious as yours’.

Marilyn

Having a sense of humour and a positive attitude were important facilitators for coping with life after stroke. Being able to achieve a positive outlook was associated with re-evaluating their lives. After her stroke, Evelyn became depressed, but through re-evaluating her situation and realising she had something to live for, she coped better. There was also a sense that a stroke survivor’s outlook was influenced by their personality prior to the stroke. For example, Denise described herself as determined and expressed that she is not one for ‘moping around’. She acknowledged that another stroke could happen, yet she is willing to take a risk and actively continue with her life rather than passively sitting at home:

So what do you think’s helped you do so well since the stroke?

I don’t know really, probably because I’m just determined to get on with things, I’m not one for sitting around moping so . . . And I’m sure I could have gone other way if, you know, I could have been getting up and going out and doing things, then I could have had another, I’m not doubting that but I’d still rather that chance and get on with life than sit around waiting for it.

Not dwelling on the stroke was another strategy adopted by some of those who were 9–12 months post stroke, for example Timmy, Denise, Jane and Evelyn. Jane did not feel the need to look back in terms of the stroke because she knew how lucky she had been. By taking this approach, these stroke survivors saw it as a way of moving forward and not focusing too much on the event itself. Although this is a positive approach to managing for some of the stroke survivors, others struggled to project too far into the future, especially when their circumstances were more unpredictable. This unpredictability occurred among those who were more impaired by the stroke and faced uncertainty in the extent of their recovery. Taking each day as it comes was a way of managing this uncertainty to some extent. Paddy has experienced three strokes since 2007; since having these, he no longer plans or thinks about the future.

Acceptance was another strategy that stroke survivors and their caregivers employed to cope with the stroke and the changes to their lives. Those who were at least 24 months post stroke talked about this more during the interviews than those who were 9–12 months post stroke. Acceptance was discussed in the context of different aspects of their recovery and in terms of coping overall with the major life event. Cedric and his wife recognised the importance of acceptance and took the view that this is something you have to ‘learn’ to do as part of a process over time:

Oh you’ve to accept it, you’ve to have a positive outlook, and you have to learn to accept it.

Cedric’s wife

Being unable to reach acceptance can be a barrier to managing and coping. Carla was 3 years post stroke at the time of the interview, but she openly stated ‘I still haven’t accepted it’. She made a distinction between realisation and acceptance:

And when do you think you started to sort of realise things would be different?

When did I realise or when did I accept?

Yeah, well whichever sort of suits?

I realised just how little I, not how little, no, how different it was all, what, how different my life is going to be when I didn’t succeed at going back to work. Because work’s a great healer, int it? You go to work you can forget it, can’t you? But no.

Failing to successfully return to work allowed her to realise that her life was going to be different, but she still struggled to accept it.

Gaining control

During the time that stroke survivors spend in hospital, control is in the hands of medical professionals, and stroke survivors become passive agents of information and support. Caregivers have not always fully taken on their roles and much of the responsibility for the stroke survivor is out of their hands. The transition from hospital to home can lead to a loss of control for both the stroke survivor and their caregivers (or other family members) through having to manage the uncertainties that living with life after stroke brings. The key areas discussed among the stroke survivors and their caregivers around control included the need to feel able to manage and the need to address uncertainties (e.g. fear of the stroke happening again and understanding why the stroke happened).

With regards to managing, there was often a mismatch between stroke survivors and caregivers in the transitional period from hospital to home. For example, stroke survivor Bob wanted control by coming home from hospital, but his wife felt that this transition would lead to a loss of control. This is when she would fully embark on her caregiving role and she was unprepared at this stage:

Again, I suppose it’s the control situation, isn’t it, of whichever side you’re of, you wanted to be in control in coming home and I wanted somebody else to be in control because I didn’t feel in control.

Bob’s wife

Both stroke survivors and caregivers needed to feel a sense of control; however, there were differences in what allowed them to feel in control. It was common for both caregivers and stroke survivors to feel abandoned and dismissed once they left hospital because of the lack of support. Marilyn’s husband would have liked someone to visit following her stroke to provide reassurance about managing initially:

You’re guessing a bit on what you’re doing, whereas you feel you need some reassurance to, for somebody to either say ‘no, don’t do that’ or ‘no, that’s fine, you keep doing that’, isn’t it?

Marilyn’s husband

Seeking and building support networks helped some of the caregivers and stroke survivors to manage once the formal support had diminished. Engaging with this support helped them to cope and feel more in control. Charlie’s wife sought informal support from their daughter when she struggled with her situation:

I get very stressed, I get very upset, I get very emotional, my daughter gets it all [laughs].

Charlie’s wife

The lack of control experienced by stroke survivors and their caregivers was also driven by the possibility of another stroke. Those who were 9–12 months post stroke still expressed fears about another stroke:

Yeah, she sounds nice. So do you worry about having another stroke?

Yeah, I do. What causes strokes?

However, this expression of fear was less common among those who were > 24 months post stroke. The stroke survivors were often unable to gain an accurate prognosis about the likelihood of re-occurrence, which added to their uncertainties. To attempt to gain control over their unpredictable situations (particularly amongst those who were 9–12 months post stroke), stroke survivors established their own cause to the stroke. Some looked to external events to explain why it happened and others considered the possibility of lifestyle choices. When possible, some of the stroke survivors actively changed their health behaviours. For example, Evelyn consciously cut down her drinking following the stroke in an attempt to reduce the chances of this happening again.

Managing emotions

The degree of emotional impact for the stroke survivors and their caregivers varied greatly. Some were really struggling, whereas others felt that the stroke had left them with no emotional impacts at all. Some of the emotions felt by the stroke survivors included shock, fear, anger, worry, anxiety, stress, depression and guilt. Some of the caregivers also experienced many of these emotions, particularly stress and strain as a result of taking on their new role.

A distinction can be made between emotional responses to the actual event and emotional responses to living with and managing the impairments and impacts of the stroke. There was a need to be able to find ways of managing, as, for many, these emotional responses were ongoing and still apparent up to 3 years post stroke. Examples of the emotions experienced by stroke survivors and their caregivers are discussed in the following distinct categories, in addition to the barriers to and facilitators of addressing these.

Doing everyday tasks around the house

Many stroke survivors (e.g. Paddy and Gordon) spent more time in their homes than they did before their stroke, as a result of being physically impaired or becoming resigned to this way of living. It was common for stroke survivors to be motivated by the need to function to a point where they could carry out purposeful tasks, for example making food or getting dressed or washed. For some ‘house-proud’ individuals, being able to carry out tasks was important for maintaining a well-kept home environment. For others, this was important as it formed part of their daily routines, which often centred on home-based activities.

Building a support network was important in cases where individuals were unable to carry out household tasks independently (e.g. cooking, vacuuming, washing). Timmy, for example, had support from his daughter when he first had his stroke as a way of coping with the demands of running his home. This was appropriate for him initially and eventually he was able to manage on his own, but just did things at a steadier pace:

So can you remember what you were doing on a typical day when you first got home from hospital?

Well me daughter stopped with me for a few days to make sure I could do things, and I was doing things gradually. Well, like I said, I had them, well ready meals just you put in the microwave, so I just had them to do, you see, but coming from the oven on this side onto the breakfast was enough.

Building a sense of togetherness was also important for some of the stroke survivors and caregivers. For example, stroke survivor Lisa and her partner worked together in managing the home environment since she had become physically disabled by the stroke:

. . . So I do the cleaning up and stuff like that, anything that needs polishing . . . or washing floors down, I do all that, then Lisa will do like surfaces that can be done while you’re stood, Lisa will do those, so yeah, yeah Lisa does what she can and then yeah I just do the other stuff.

Lisa’s husband

Encouragement from a partner was also useful for some of the stroke survivors as they managed tasks around the home. Jimmy’s wife let him try tasks before stepping in to support him. Confidence was also influential in whether or not some stroke survivors could carry out everyday tasks around the house. Cathy did cooking in hospital before she was allowed home. She perceived this as an exam that gave her the confidence to feel prepared for her return home. Repeatedly carrying out tasks without assistance increased her confidence:

Was there certain things you did to prepare or were you just ready . . .?

I said, ‘once I’ve passed those exams in hospital, I’ll be alright’, you know and then they put me in a room, quite a big room that was carpeted for like home-from-home, so you’re on your own and you have to do everything yourself, so when I’d done that for about 4 weeks they thought, ‘oh she must be OK’.

Some stroke survivors talked about acceptance in the context of knowing that jobs around the house take longer and they commonly managed daily tasks by pacing or creatively problem-solving. It was also common among the stroke survivors to start to see themselves as agents in their adjustment and recovery. This involved taking some responsibility for their actions and recovery process. For example, Paddy purposefully uses his weak arm in tasks of daily living (e.g. shaving) to be able to manage day to day and facilitate improvements.

Caregivers being overprotective was sometimes a barrier to stroke survivors carrying out tasks around the home. Caregivers were inclined to be most protective in ‘risky’ areas around the home, for example the kitchen and the bathroom. These were places they associated with greater risks of falls or injuries.

Working towards physical/functional improvement

Many stroke survivors in the sample were physically affected by the stroke to some extent. Some (e.g. Cathy) were unable to walk without a walking aid. Others experienced numbness in their limbs or loss of function/co-ordination (e.g. Robert). Some had no physical impairment at all (e.g. Jane). Stroke survivors who had been affected physically by the stroke demonstrated a need to continue improving once the formal rehabilitation ended. These improvements were often recognised in the context of their abilities to carry out purposeful tasks or meaningful activities and were important for gaining some normality.

A number of facilitators of improvement were identified among the stroke survivors. Creative problem-solving helped enable physical functioning, despite physical difficulties. Marilyn and her husband had an exercise bike that she was unable to use using conventional methods, but her husband found a beneficial way for her to adjust her technique, so she could do her exercises:

Well, because what we do is, we turn it the other way round, ’cause Marilyn can’t sit on it, so I turn it with the handlebars facing her and then she can pedal it sat in her chair the other way round.

Marilyn’s husband

Some stroke survivors also talked about the encouragement they gained from significant others/partners/friends/peers. Some caregivers suggested exercises for the stroke survivors, and prompted movement either when formal therapy ended or when this was never provided:

My husband’s very good, like I said, he thought of the ball idea and he’d say to me, ‘lift your leg, lift your leg’, every so often, ‘lift your leg, keep doing that’, you know, and ‘move your foot, bend your foot like that and that, push it up and down’ so many times a day . . .

Mavis

Among those at 9–12 months post stroke, hope was important, particularly when improvements were uncertain:

Oh, I hope so, yeah. I hope my legs will get working better, that’s the only thing to go now.

Cathy

It was not as common for those who were at least 24 months post stroke to remain hopeful because they had managed for longer and come to terms with the extent of improvement in their recovery. These individuals may have reached a new sense of normal at this stage with regards to their physical functioning.

Motivation was an important facilitator for physical improvements. Those who were highly motivated recognised improvements and emphasised the importance of continued improvement be able to carry out meaningful tasks. On the other hand, difficulties recognising improvements were a barrier to continuing to work towards improvements. If stroke survivors failed to recognise improvements, they became disheartened and lacked motivation. For example, Marilyn could not see that she had improved, despite her husband telling her. She needed reassurance from a professional to be able to see that she was doing ‘alright’:

I’m alright once I went to physio[therapy] and they say ‘yes you are’, you know, ‘you’re doing alright’, if they say I was doing alright, I mean, I felt totally different then about it all.

Marilyn

Unfortunately, the input Marilyn received from the physiotherapy team ended, which made this situation more difficult, as her husband could do little to persuade her that she was improving. This is closely linked to another barrier: the perceived lack of physical rehabilitation support. Some stroke survivors thought that they could have benefited from more input from physiotherapists to facilitate a better physical recovery.

Managing hidden consequences of stroke

Many stroke survivors experienced consequences of the stroke that could be considered more ‘hidden’, for example difficulties with concentration and processing of information, memory impairments and mood swings. These were things that were often hidden from caregivers, the public and health professionals. They were also sometimes hidden from the stroke survivors if they were unable to recognise changes in themselves, for example a change in temperament. Stroke survivors needed to find ways to manage these hidden consequences.

Creative problem-solving facilitated the management of hidden consequences. For example, Denise found it useful to write things down to cope with the demands of her job at the opticians. By using this strategy, she did not feel that her memory loss interfered with her working duties:

So it doesn’t interfere with anything at work, like, it’s not a problem or anything if you don’t remember something?

No, because I have a little tray and I write things down and put notes in me little tray.

Acceptance was important for coming to terms with the longer-term consequences of stroke, such as difficulties processing information, and the resultant impacts that these can have on their lives. Driving is an example of an activity that many stroke survivors struggled with as a result of both physical and cognitive impairments. Carla experienced ‘hidden’ consequences, for example problems with spatial awareness and processing of information, that contributed to her decision to stop driving. Her doctor helped her to realise that she must be truthful and face her circumstances following the stroke. She accepted that she could not deal with the physical or mental side of driving, although she still misses this aspect of her pre-stroke life.

Roger also experienced difficulties with his ability to process information. Building a support network was a way in which he and other stroke survivors have managed this problem. Rogers’s wife wrote instructions for him for when he is using the printer. This worked to some extent, but he often put the instruction card away, meaning his wife had to prompt him to get this out:

I’ve written a little notice by the printer, telling him which order to do everything, but he puts the card away. [Laughs] I said if I write it out I’ll know which way to do it but it’s alright, when we need I say, ‘shall we get the card out?’, you know.

Roger’s wife

Although there are positive ways in which some of the stroke survivors have managed, they also faced barriers to managing at times, one of which was difficulty recognising changes as a result of the stroke. Roger has also had impaired concentration since the stroke, yet he does not recognise this. He and his wife experience conflict around his driving abilities, as he thinks that, if he was physically fine, he would be able to drive. His wife has noticed his lack of concentration, yet this is hidden from him. This makes things difficult because he is unrealistic about what he can still do:

Mm, so for you it’s all about getting your feet working?

Feet, yes. I mean if a car was the opposite way round I could drive it, I’m sure I could.

It’s the concentration though Roger, you don’t see things quite the same way.

Well, you and I disagree on this one.

Experiencing a lack of control over difficulties such as mood swings and irritability can also serve as a barrier to managing these issues. For example, Carla experienced these problems for some time following her stroke. She often became very irritable but felt that she could not stop herself. Over time, she is more aware of these difficulties and can control these changes in mood more easily than she did initially. She can keep her emotions in check, but it doesn’t happen as much as it used to. This is something she talked about as a past behaviour and on reflection described this as ‘horrid.’

Stigma was also problematic in managing the hidden consequences of stroke. Lizzy experienced ‘public stigma’ as she found that people struggle to see that she is ill and associated her stammer from the stroke with being drunk:

Well it’s this problem that people can’t see that you’re ill, you know, they don’t realise that, it’s a bit like being deaf, they can’t see you’re deaf, they can’t see your speech is trouble, so they don’t know whether you’re, whether you’ve got a stammer or whether you’re drunk!

Lizzy

When she is out in public, she prefers that people know what her problems are, so that they cannot make these assumptions.

Obtaining usable information

There was a general sense of negativity attached to the timing and content of the information that was given to stroke survivors and their caregivers in hospital. Despite this, stroke survivors and caregivers expressed a need to obtain usable information. The extent of usability was dependent on different circumstances, preferences and needs. Some of the stroke survivors expressed a preference for more specific information to meet needs such as managing incontinence and making lifestyle and dietary changes. Some also wanted to be more informed about what to expect following the stroke.

An example of a common barrier to finding the information usable initially was ‘information overload’. Some of the stroke survivors felt that they had been bombarded with information to the point where they failed to look at it:

Oh yeah, you get all sorts of information, leaflets you know, but you don’t tend to take it all in because you get that much of it, you know.

Malcolm

Others found the information unsuitable because it was specific, too general or too obvious.

Timing of information was also problematic for some stroke survivors and their caregivers. Lisa and her husband felt that the timing of information from the speech and language therapist was poor because they had already devised their own communication methods by this point. Her husband would have liked some advice while she was in hospital on how to communicate with her, and found that the cards given to them later by the speech and language therapist would have been more appropriate at that stage:

. . . she brought the charts with her, the letter charts so that and then she brought charts that had pictures and stuff so you could, so Lisa could just be shown at it, so she could just point at things, like I say I’m hungry or I want a cup of tea, so I looked at them and I thought, well they’d have been handier in the hospital.

Lisa’s husband

Once they got the communication aids, they could have benefited from some more training, as Lisa struggled with getting used to the tablet and they were offered little choice.

Actively seeking information via asking questions or using the internet was a facilitator of ensuring that information was more usable. This was common when stroke survivors or their caregivers faced a problem and wanted to learn more to address the problem. Maude’s husband used the internet to gain information and reassurance about his wife’s diagnosis and prognosis of stroke while she was in hospital:

Well just look at the different types of stroke and the, and the difference between, you know, what the symptoms are of a transient ischemic attack and a full stroke and things of that nature, and the different types of stroke, whether they’re haemorrhagic or not and that was what I looked at, and it did seem that if you had to have a stroke, the best sort of stroke to have would be the one that Maude had so.

Maude’s husband

Building a support network is closely linked to actively seeking information. This enabled some of the stroke survivors and their caregivers to obtain the types of information they needed from a certain source of support. Jimmy and his wife attended a group for carers where they took part in a course over a number of weeks. They had the opportunity to engage in a one-to-one session at the end of the course to gain information that they may not have wanted to discuss in front of the wider group:

By the end of the course you could go up to them and say, you know, a one-to-one with anyone and say, ‘is there . . .?’, you know, and if you’ve got something that you’re thinking about you might wanted to have keep it private you know, between the person, not let anybody else on the course know about it, but you can speak to her, you know?

Jimmy

The level of interest and extent to which it meets their needs also plays a part in whether or not some of the stroke survivors make use of information. Alfred subscribed to a monthly magazine regarding his health and found it interesting to see how different people deal with their strokes. Cathy used information she was given in a folder by one of the community nurses. This comprised numerous sections, examples being ‘life after stroke’ and ‘social care and support.’ This information met her needs in terms of providing her with some understanding about the stroke and useful contacts for managing specific issues, for example benefits.

Sustaining flexible support networks

Stroke survivors used both formal and informal support networks as part of managing their ‘recovery’. Most of the stroke survivors talked positively about the formal support they received from health-care professionals while they were in hospital. Following discharge from hospital, there was a shift in the types of support they received. Examples of post-discharge support included family, friends, community nurses, other external agencies (e.g. Age UK, Stroke Association) and support groups. Once the stroke survivors left hospital, they had more choice in the support they wished to mobilise, although they were still restricted to some extent by access and availability. Support was also influenced by changes in circumstances of support providers (e.g. friends and family). Stroke survivors and their caregivers needed to find ways of sustaining support in the event of unexpected changes.

Stroke survivors and their caregivers experienced a range of barriers to being able to sustain support. Formal support such as physiotherapy was not available to all stroke survivors; this left some (e.g. Mavis) feeling angry because they wished they could have had this support. Many stroke survivors also suffered from a lack of support as, sadly, many family members and friends had died. Lizzy lost her husband, who died of cancer very close to the time of her stroke; this made managing with the stroke more difficult because he wasn’t around to help her. She described this as a ‘double whammy’ because she had to cope with being widowed and manage the impacts of the stroke:

Yeah. It was the double whammy of having, being widowed and the stroke so close together that it’s taken a bit longer. They did suggest that I’d had the stroke was maybe because of I hadn’t grieved properly for my husband.

Lizzy

Acceptance (or not) of support was one of the facilitators of being able to sustain support. Stroke survivors and their caregivers talked very little about support from social care and talked mostly about health care, third-sector organisations and informal support. Individuals were often accepting of support from friends, caregivers and relatives, yet more formal, group-based support was not as readily accepted. The extent to which support was accepted related to how appropriate it was for addressing their needs. David and Jane were not interested in attending groups because they did not think their strokes were bad enough.

As with the information, actively seeking support is a facilitator of gaining useful flexible networks. Lizzy had some input from a speech and language therapist, but did not feel that she was getting on very well, so she proactively got in touch with her to overcome this problem. She was given more exercises to try following this:

I called her back about 4 months after because I didn’t feel that I was getting on very well and she gave me some more exercises.

Lizzy

Accessibility can be both a barrier to and a facilitator of being able to engage with some types of support. Some stroke survivors wanted to be able to attend groups, but they were limited by their location and transport options. Others could be visited at home for health-related support, in which case accessibility was not problematic.

Engaging in meaningful activity

Stroke survivors needed to engage in meaningful leisure and home-based activities including golf, board games and housework. The meaning attached to these activities varied for different individuals. Some stroke survivors valued the social aspects of activities, others engaged in activities for enjoyment, to maintain a purpose or to keep busy.

Some individuals could engage in the same activities as before their stroke, some took part in some aspects of previous activities and others changed their activities as a result of the impairments caused by the stroke. For example, before her stroke Cathy was a teaching assistant. She was unable to continue in this role because of her physical and communication difficulties, but she had more time to engage in housework. She found new enjoyment in her role around the home, which keeps her busy following the stroke:

Oh, I’m just as busy in the house; I never thought I’d like housework! So I do that, I’ve been in the cloakroom with a vacuum this morning and I’ve got a bread machine so I make bread for us.

Cathy

Engagement in previously enjoyable activities was influenced by their physical abilities and level of confidence. Daphne played bowls before her stroke and lacked confidence initially. However, her captain encouraged her to keep playing, which gave her the confidence to continue, despite some physical impairments post stroke.

Accessibility was also influential in whether or not stroke survivors could continue to engage in meaningful activities. Some stroke survivors had informal support networks that could provide transport to enjoyable activities, which facilitated ongoing engagement post stroke. However, not all stroke survivors could rely on such support. Carla wanted to be able to attend a group, but, because of her rural location, lack of family close by, limited bus routes and inability to drive, her access was limited.

For some, the level of engagement in new or previous activities was influenced by motivation. Some stroke survivors lacked motivation following their stroke, which meant that they were reluctant to engage in activities. A change in or loss of identity as a result of physical changes to the self also served as a barrier to engagement in previously enjoyable activities. Paddy became physically impaired following his strokes, leading to changes in his self-concept. He described himself as active before the stroke and this loss of self has been a bit of a ‘comedown’ for him. To meet the needs of his new defined self as ‘less active’, he changed his activities, meaning he spends prolonged periods watching rather than playing football.

Some stroke survivors benefited from being encouraged to take on new activities. Cedric’s wife bought him a computer because she believed that it would be useful as part of his cognitive recovery. He later joined a computer class, which he finds enjoyable. Linked to this was the importance of building a support network for engaging in meaningful activities. This included support in a social sense or for physically attending an event. Lizzy is a representative of a sugarcraft club. She finds people who she can ‘double up’ with to get to her regular meetings, as she has lost her confidence with driving longer distances:

Yes. I’m a rep[resentative] for the sugar people as well, sugarcraft people and if we’re having meetings down in Brighton or something like that, I usually try and find someone I can double up with.

Lizzy

It was more difficult for stroke survivors to continue with activities if they experienced a lack of interest. Alfred no longer plays table tennis since he had his stroke; finding a potential replacement for this activity has been difficult, as table tennis provided him with enjoyment. It is the physical aspect of this activity that he has struggled to replace and he lacks interest in alternative activities such as bowls. Lack of money can also have an impact on the level of activity that stroke survivors can engage in. Carla lost her job as a nurse as a result of the stroke. She was caught up in a vicious circle as she is limited in what she can do without money, yet she was struggling to figure out what job she would be able to do instead.

Overcoming financial concerns

Some of the stroke survivors and their caregivers talked about financial difficulties during the interviews. Some stroke survivors and their caregivers had to give up work following the stroke and others were forced to increase their working hours to be able to cope financially (e.g. Paddy’s wife).

Obtaining benefits allowed some of the stroke survivors and their caregivers to manage, but the process for obtaining these was often problematic. Therefore, stroke survivors and caregivers had a need to find ways of overcoming this financial concern.

A common barrier to being able to obtain benefits was lack of information or awareness:

He said ‘I’m not being funny but you’ll get £200 a month for what you are’. ‘Oh’, I said ‘I wouldn’t know where to start or where to go’.

Timmy

The facilitator of overcoming this issue was ‘knowing the process’. Cathy emphasised the importance of knowing about the process of getting benefits earlier on. She knew about this only because she was given an information pack, but she found it stressful managing financial issues alongside adjusting to the impact of the stroke:

I think definitely people need to know about and they should be told that, I think, quite early on because I think it was November or no, October, I got in touch with the Independent Living Allowance and they send the form and it had to be in and again they can take as long as they want and you have a deadline, you have to have it with them.

Cathy

Building a support network was a facilitator of overcoming some of the issues around benefits. Some of the stroke survivors made links with charities such as Age UK that provide financial support (e.g. Lisa and her husband).

Maintaining relationships

There was a need among stroke survivors to be able to maintain relationships with significant others, including partners, other family members, colleagues and friends. The interview accounts reflect the diverse effects that a stroke can have on relationships. Some of the stroke survivors and caregivers in the sample experienced difficulties, but did not face a complete breakdown in their relationships as a result of the stroke. For example, Charlie’s wife recognised that he has become more ‘chompy’ since the stroke, and described her new life as a caregiver as stressful. Alternatively, others have become closer since the stroke. For example, Lisa’s husband believes marriage is about caring for each other so he takes the view that there is a change in that care. He does not think of himself as her carer and understands that she would do the same if the situation was reversed. This supports the importance of maintaining relationships for both the stroke survivors and the caregivers:

You don’t think of yourself as a carer?

No, it’s just one of those things, when you’re a married couple you care for each other anyway, all it is is just a little bit of change in that care, that’s all, for me that’s all it is, it’s that the only . . .

The thing that’s different, yeah, it’s a nice way of looking at it.

If it were the other way around, I know damn well she’d be exactly the same, you know what I mean, if things ever changed, if she had to do things that I’d normally do, she’d do it, so just that, just how it should be, isn’t it?

The concept of togetherness facilitated a maintenance of relationships. For Lisa’s husband, the sense of togetherness is reflected in his understanding that he is there for his wife and he knows this would be reciprocated if the situation was different. Cedric and his wife similarly talk about ‘looking after each other’, yet they talk about this in terms of how they work together as a team on a daily basis. There is a sense of appreciation and understanding, which has allowed them to maintain a positive relationship following Cedric’s stroke.

Acceptance also played a significant part in the maintenance of relationships. Some stroke survivors talked about this as a process that they worked towards over time. Paddy provided an account of how he reached acceptance. This involved consciously making a decision to accept the stroke because he did not want to become bitter and thinks he would have driven people away had he not accepted it:

. . . I think you’ve got to accept it, otherwise you’ll be snapping at everyone and falling out with everyone. You drive everyone away. You’d end up with no friends because people would get fed up of you, wouldn’t they?

Paddy

‘Building a support network’ of informal support was another approach to maintaining relationships with family members and friends beyond the relationships between the stroke survivors and their partners. Malcolm’s wife passed away 4 years prior to the stroke. He continues to maintain relationships with his friends, who have been an important part of his life since he has lived alone.

‘Disrupted couplehood’ is one of the barriers to maintaining a positive relationship. Conflict is one of the factors that can lead to disruption in relationships. Roger and his wife had disagreements regarding driving. He has struggled to come to terms with the loss of driving and, according to his wife, has become a back-seat driver. She believes this has caused a few problems and throughout the interview it was evident that they have had a number of arguments since Roger’s stroke:

No, but there’s still . . . But the one thing with the driving is, Roger so misses the driving and my driving according to Roger is hopeless so . . . [Laughs].

Yeah, mine is hopeless according to my husband as well.

So it’s like having a back-seat, a side-seat and a front-seat driver! [Laughs] And it does cause a few problems.

‘Couplehood’ can also be disrupted by the loss of sexual relationships. Charlie and his wife recognised that their sexual relationship had suffered following the stroke. Paddy also mentioned that the stroke has affected the sex life of him and his partner, although he stated that they have both accepted this. His partner was not present in the interview; therefore, her perspective on this remains unknown.

The ‘burden of caring’ can also serve as a barrier to maintaining the relationship that some of the stroke survivors had with their partners before the stroke. In terms of managing each day, the stroke had an impact on some of the partners. When Lisa is feeling down, it also knocks her husband down. Charlie’s wife also feels the stress in coping with life after stroke. She thinks he can be very demanding now and it can be difficult making sure he is OK:

Yeah, how is it stressful? What are the most stressful things about it?

Well it’s making ends meet and still running the house and making sure he’s OK and Charlie can be demanding at times, can’t you, love? When you’re in, when he’s having a bad day he can be . . .

Despite their caregivers feeling this burden at times, Charlie and Lisa have managed to maintain relationships, albeit that the dynamic has changed since the stroke. It is clear that the impact of the stroke can lead to a change in roles. This change was experienced more positively by some than others. This can be problematic for a relationship when the partners are no longer happy with their new role. For example, Michael’s wife took on a new gardening role, which would not be her preferred choice prior to the stroke. In such circumstances, it was important for couples to work together to try and embrace their new roles.

Becoming a caregiver is one of the most significant role changes. Their role typically involves spending most of their time looking after their loved one, who is dependent on them for support every day. Maintaining relationships can become difficult when there is a negative perception of the resultant change in dynamics. However, this new dynamic in the relationship can be positive when aspects of their marital relationships can be maintained. Arnie and his wife still chat together and have a cuddle as they would have done before the stroke.

Reconstruction of identity

A loss of functioning in part of the body or loss of confidence or memory following the stroke can lead to a change in a stroke survivor’s sense of self. Paddy experienced a loss of identity after the first of his three strokes left him physically impaired down one side. His sense of self as an active person has been lost since he has been unable to engage in physical activity. During the interview, he described his new self as ‘restricted.’ A reconstruction of identity is often required to address the loss of or change in identity as part of adjustment to life after stroke. Some require little reconstruction if they can maintain aspects of their previous self, for example independence. For others this can be more difficult and it means adjusting to take on new self-concepts that they did not previously hold.

Stroke survivors expressed the barriers to and facilitators of reaching the construction of a new identity. Acceptance was an important facilitator of achieving this. Iris knew that she was going to be disabled by her impaired functioning in her arm and leg, but she did not feel that there was anything that she could do, so saw no point in getting worried.

Self-concepts can be related to previous activities. It was therefore important for stroke survivors to find meaningful activities that suited their new identities, or allowed their previous identities to remain. Iris refers to herself as disabled following the stroke; when she was asked about herself as a person, she talked about herself in terms of the physical disability she has acquired. As she has adjusted to this new identity, she has looked for ways of meeting her needs, such as a holiday for people with a disability:

I have but I’ve never, I’ve said, I’ve told you about holiday flats or houses or that, well that’s not what I want, it’s something where I can go as an individual, that I could be taken there and sort of looked after there, carefully looked after and then brought back as well, but who understands what it is to be disabled.

Iris

Irreplaceable loss was a barrier to managing the new identity following the stroke. This refers to loss of what allowed them to define themselves as in a certain way prior to the stroke (e.g. as an independent person) and is related to different aspects of their lives (e.g. driving and working). Carla valued her independence in terms of being able to do what she wanted. Following the stroke, she has experienced a sense of irreplaceable loss in terms of this aspect of her identity. She acknowledged that she will not be what she was before (physically) and finds this really hard.

Stigma was also challenging for stroke survivors as they worked through a process of reconstructing their identities. During the interviews, stroke survivors talked about stigma in relation to being old, being like a baby, looking drunk and being less intelligent. Cathy perceived herself as intelligent before she had her stroke. The stroke caused her to become physically impaired and she needs to use a wheelchair when she goes out beyond her home. Visibly, she entered the social world with a new identity, but her intelligence was unaffected by the stroke. She noticed people acting differently towards when she used the wheelchair:

Well, just that they talk down to you. They don’t think you’re as intelligent and it’s not affected my intelligence.

Cathy

Fortunately, this did not stop Cathy from going out, but it could be problematic for other stroke survivors.

Managing beyond the home

Managing with life after stroke goes beyond the home environment. The home can be a safe space for regaining some of the control that is lost at the time of the stroke. In contrast, stroke survivors and their caregivers can feel less in control beyond their own home, as the outside world leads to numerous challenges. These include those that are within the built environment (e.g. inappropriate access for disabled people and lack of handrails) and those that occur in interactions with others. It is therefore important for stroke survivors to be able to manage beyond the home environment.

Cedric and his wife provide an example of a situation that reflects the difficulties that stroke survivors can experience beyond the home, both physically and in interactions with others. The built environment meant that the wheelchair became stuck at the airport, leading him to fall. People did not know what to do, so he was left struggling in this situation. Cedric tried to account for why people did not help him and suggests that this is a result of a fear of being sued. He also talked about lack of public awareness in terms of handling a stroke survivor, which complicated matters further for him:

And he’s, he’s difficult to get up now; for instance, yesterday, when we were coming off the plane, we always have a wheelchair for Cedric and there was a young girl, like, you, god love her, I felt right sorry for her, pushing Cedric, and there was a kind of a little hump and didn’t the wheels get stuck and he fell straight out of the wheelchair, right out on his legs, ‘cause he can’t save himself. So you know, I’m saying, ‘Don’t panic now, don’t panic,’ and I’m worse than anybody does, so once I can get him around he can get on his knees, then I can help him up, you know, and that’s a . . . you know, there was people at the back of us and they just stood looking at us.

People don’t want to get involved because they’re that afraid, I think, of being sued.

Yeah, that’s true.

We’re not that kind of people, I took them into Morrison’s [Morrison’s Supermarkets plc, Bradford, UK] and he had a hot cup of tea.

Yeah, but they don’t know, a lot of people how to lift a person that’s had a stroke, you know, they get hold of your hand and try and pull it, well you don’t, you know what I mean [laughs]?

Examples of barriers to and facilitators of managing beyond the home are discussed as follows.

Accessibility can be both a barrier to and facilitator of managing beyond the home. Iris has no problem getting to places, as she makes use of taxis. However, she struggles when she arrives at her destinations. Therefore, she is reliant on support from others to manage. This links to the importance of ‘building support networks’, a facilitator for managing beyond the home. Stroke survivors often needed such support to feel safe:

Yeah, that’s what me daughter says you see, I sometimes take her out shopping and that on a weekend and she says, ‘You’re looking for your trolley when you get out, “Where’s the trolley?”. You don’t need your trolley, you’ve got me here’, which is right because, should something go wrong, she’s there.

Timmy

Walking aids were also a facilitator of physically managing for some of the stroke survivors (e.g. Timmy, Cathy). This was another way of protecting their safety and facilitating movement. Timmy has his trolley as a walking aid when he walks longer distances alone. This stops him from panicking in the absence of support.

Some barriers were also apparent for managing beyond the home where stroke survivors faced interactions with others. Some stroke survivors were reluctant to use wheelchairs and sticks because of the stigma attached to these. As a result, they did not often go outdoors. Paddy has limited mobility following his stroke. He uses a stick around the home and for short distances beyond the home. However, he is reluctant to use his wheelchair for longer distances because he thinks he looks like a baby if his wife pushes him around:

If she’s pushing me about, it’s like being a baby in a pushchair.

Paddy

Paddy assumes that people will think in this way; he has never had an experience where someone has told him he looks like a baby. He admitted he could be wrong and he might just be ‘being silly’, but it still stops him from going to some places, which affects the activities he can engage in.